Acta Crystallographica Section D Biological Crystallography
ISSN0907-4449Crystallization and X-ray diffraction analysis of insect-cell-derived IL-22
Ting Xu,a,b Naomi J.Logsdon a and Mark R.Walter a,b*
a Center for Biophysical Sciences and Engineering,University of Alabama at Birmingham,Birmingham,AL35294,USA,and
b Department of Microbiology,University of Alabama at Birmingham,Birmingham,
AL35294,USA
Correspondence e-mail:walter@uab.edu
#2004International Union of Crystallography Printed in Denmark±all rights reserved Interleukin-22is a potent mediator of cellular in¯ammatory
responses.Crystals of interleukin-22expressed in Drosophila
melanogaster S2cells(IL-22Dm)have been grown from polyethylene
glycol solutions.To obtain crystals suitable for X-ray diffraction
analysis required the separation of different IL-22Dm glycosylation
variants and the inclusion of the detergent cetyltrimethylammonium
bromide(CTAB)in the crystallization experiments.The crystals
belong to space group P21,with unit-cell parameters a=64.88,
b=62.23,c=139.524AÊ, =91.35 ,and diffract X-rays to2.6AÊ
resolution.The crystallographic asymmetric unit contains six
IL-22Dm molecules,corresponding to a solvent content of approxi-
mately49%.
Received1March2004
Accepted29April2004
1.Introduction
Interleukin-22(IL-22)is a recently identi®ed
-helical cytokine produced by activated T
cells(Dumoutier,Louahed et al.,2000;Xie et
al.,2000).The biological function of IL-22is
still being de®ned.However,several studies
show that it plays an important role in
in¯ammatory responses.In particular,IL-22
has been shown to up-regulate the production
of early systemically circulated defense
proteins(acute phase proteins)such as serum
amyloid A, 1-chymotrypsin and haptoglobin
in liver cells(Dumoutier,Van Roost et al.,
2000).IL-22also up-regulates gene expression
of pancreatitis-associated protein,PAP1,in
acinar cells(Aggarwal et al.,2001)and induces
the production of reactive oxygen species
(ROS)in B-cells(Wei et al.,2003).IL-22
biological responses are mediated through a
heterodimeric receptor complex consisting of
the cell-surface receptor chains IL-22R1and
IL-10R2(Dumoutier,Louahed et al.,2000;
Dumoutier,Van Roost et al.,2000;Kotenko et
al.,2001;Xie et al.,2000).
IL-22is a member of the class-2cytokine
family,which includes IL-10,IL-19,IL-20,
IL-24and IL-26(Fickenscher et al.,2002;
Pestka et al.,2004;Walter,2002).Class-2
cytokines share$25%sequence identity and
have a largely conserved IL-10footprint
sequence(Walter&Nagabhushan,1995).The
crystal structure of IL-22produced in Escher-
ichia coli(IL-22Ec)has been determined
(Nagem et al.,2002).IL-22Ec exists as an
-helical monomer that is structurally similar
to IL-10(Logsdon et al.,2002).However,in
contrast to IL-10,IL-22expressed in eukary-
otic cells is glycosylated at one or more of its
three N-linked glycosylation sites[Asn-X-Ser/
Thr(N X S/T),where X is any amino acid].The
IL-22Ec crystal structure reveals that the three
asparagines(Asn54,Asn68and Asn97)
involved in N X S/T glycosylation sites are
located on helix A,the AB loop and helix C of
the structure.
To test the role of glycosylation in receptor
binding,IL-22was expressed in Drosphila
melanogaster S2cells(IL-22Dm)that attach
hexasaccharides to N X S/T glycosylation sites
(Manneberg et al.,1994).Surface plasmon
resonance(SPR)analysis reveals that a soluble
extracellular fragment of the IL-10R2chain
exhibits a tenfold increase in af®nity for
IL-22Dm(K d=100m M)compared with
IL-22Ec(K d91m M)or an insect-cell-
expressed IL-22mutant(K d91m M)where
the N X S/T sites have been destroyed by
mutagenesis(N.J.Logsdon&M.R.Walter,
unpublished results).Here,we report the
crystallization of IL-22Dm,which may lead to
an understanding of the structural basis for
N-linked carbohydrate in IL-10R2receptor
recognition.
2.Materials and methods
2.1.Expression and purification
Restriction enzymes were purchased from
New England Biolabs.All PCR experiments
were performed using Pfu polymerase(Stra-
tagene).An IL-22expression construct was
created by amplifying IL-22cDNA from
the plasmid hILTIF/pCEP4(Renauld).The
N-terminus of IL-22was modi®ed to code for a
six-histidine af®nity tag followed by a factor Xa
protease site.The ampli®ed PCR product was
cut and ligated into the pMT/V5-HisA
Acta Cryst.(2004).D60,1295±1298DOI:10.1107/S09074449040104921295
expression plasmid(Invitrogen)to form pAHF-IL-22.
Transfection and selection lano-gaster S2cells with pAHF-IL-22and the hygromycin-resistance vector(pCpHYGRO) were performed as suggested by the manu-facturer(Invitrogen).Expression was induced at a cell density of5Â106cells mlÀ1 by the addition of200m M Cu2SO4(0.5m M ®nal concentration)and allowed to proceed for4d.Cells were removed by centrifuga-tion and the expression medium was dialyzed into buffer consisting of20m M
Tris±HCl pH7.9,0.5M NaCl(buffer A) containing5.0m M imidazole.1l dialyzed medium was subsequently applied onto a 10ml nickel-af®nity column(Novagen).The protein was extensively washed in buffer A containing20m M imidizole and subse-quently eluted with a1M imidizole step gradient.Eluted fractions containing IL-22Dm were dialyzed overnight into 20m M Tris±HCl pH8.0,1m M EDTA, 100m M NaCl and3m M CaCl2.The His tag was removed by overnight incubation of His-tagged IL-22Dm with factor Xa [1:50(w:w)].The resulting digestion mixture was diluted tenfold with20m M Tris±HCl pH8.0and loaded onto a1.6ml Poros20HQ column.Puri®ed IL-22Dm was collected in the¯owthrough and concentrated to 10mg mlÀ1using a Centriprep10(Amicon). Protein concentrations for all experiments were determined by UV absorbance at 280nm using a calculated40.1%value of0.24.
2.2.Mass spectrometry
Native and deglycosylated IL-22Dm samples were analyzed by MALDI±TOF mass spectrometry(MS)on a Voyager Elite mass spectrometer(Perseptive Biosystems) operating in postive mode.Samples were mixed in a1:10ratio with sinapinic acid dissolved in1:1acetonitrile:0.1%TFA. Enzymatic deglycosylation experiments on IL-22Dm were performed with PNGaseF (New England Biolabs)under native conditions.IL-22Dm samples(0.5mg mlÀ1) in20m M NaCl,20m M Tris±HCl pH8.0 were incubated overnight with dilutions of PNGaseF ranging from0±500times the supplier's recommended enzyme concen-tration required for complete cutting of denatured proteins.
2.3.Crystallization
All crystallization experiments employed the hanging-drop method.Crystallization drops consisted of1m l of IL-22Dm or IL-22Dm cation-exchange fractions at 10mg mlÀ1in20m M Tris±HCl pH8.0combined with1m l of a solution consisting
of100m M trisodium citrate pH6.0,0.2M
ammonium acetate,1.4m M CTAB and20%
PEG4000.The drops were equilibrated at
310K against a well solution consisting of
0.1M trisodium citrate pH6.0,20%PEG
4000.All buffers and precipitant solutions
were warmed to310K prior to setting up the
experiments.Despite numerous attempts at
optimization,the conditions described
yielded the highest quality crystals obtain-
able for the heterogeneous IL-22Dm protein
and the cation-exchange fractions F6±F9
described in x3.1(see Figs.1and2).
3.Results and discussion
3.1.Characterization of IL-22Dm
SDS±PAGE analysis of puri®ed IL-22Dm
revealed three bands between$16and
$20kDa that are consistent with the
presence of N-linked glycosylation at each
of the three N X S/T sites(Fig.1,lane a).To
con®rm this hypothesis,IL-22Dm was
enzymatically deglycosylated with PNGaseF,
which cleaves the glycosydic bond and
converts the N-linked asparagine residue
into an aspartate.SDS±PAGE analysis of the
PNGaseF-treated samples reveals four
protein bands separated by$1kDa,corre-
sponding to IL-22Dm species with three,
two,one and no N-linked carbohydrate
species,respectively(Fig.1,lanes b±d).To
con®rm this analysis,mass spectrometry was
performed on puri®ed IL-22Dm(Fig.1,lane
a)and PNGaseF-treated samples(Fig.1,
lanes b±d).Mass-spectroscopic analysis of
puri®ed IL-22Dm(Fig.1,lane a)identi®ed
six peaks with molecular massess of19722.8
(peak1),19575.2(peak2),18684.6(peak
3),18832.9(peak4),17792.1(peak5)and
16749.0(peak6).These same peaks were
observed in different ratios in the IL-22Dm
sample treated with0.06units of PNGaseF
(Fig.1,lane b).Only peak6and an addi-
tional peak with a mass of17639.2were
observed in the IL-22Dm samples treated
with6and300units of PNGaseF(Fig.1,
lanes c and d).The147Da difference
between peaks1and2and peaks3and4
cannot be resolved by SDS±PAGE,resulting
in the four gel bands observed in Fig.1,
which correspond to the number of N-linked
glycans(0±3)attached to IL-22Dm variants.
Using the expected weights of the carbo-
hydrate moieties and the IL-22polypeptide
chain without N-linked carbohydrate
(16749Da),the glycosylation state of
IL-22Dm was further de®ned(Table1).
Previous characterization of insect-cell
glycosylation has shown the most common
N-linked glycan is a hexasaccharide with a
molecular weight of1039Da,which consists
of two N-acetyl glucosamines(GlcNAc,
203Da),three mannose residues
(Man,
reactive materials studiesFigure1
SDS±PAGE gel of IL-22Dm treated with PNGaseF.
Lanes a±d correspond to IL-22Dm protein treated
with0(a),0.06(b),6(c)and300(d)units of
PNGaseF.Gel bands are labelled according to the
number of N-linked glycans attached to IL-22.The
band at$35kDa in lane d corresponds to
PNGaseF.
Figure2
SDS±PAGE gel of cation-exchange-puri®ed IL-
22Dm glycosylation variants.Lane S corresponds to
the IL-22Dm starting material.Lanes F6±F9corre-
spond to IL-22Dm cation-exchange fractions6±9.Gel
bands are labelled according to the number of N-
linked glycans attached to IL-22.
Table1
IL-22Dm glycosylation variants identi®ed by MALDI±TOF MS.
The peak number corresponds to the qualitative estimated abundance(determined by SDS±PAGE and MS)of the variant,with1being the most abundant.
Mass
Peak No.IL-22Dm glycosylation variant No.glycans Observed Predicted 1IL-22+2-(GlcNAc2Man3Fuc)-1-(GlcNAc2Man3)319722.819719
2IL-22+1-(GlcNAc2Man3Fuc)-2-(GlcNAc2Man3)319575.219572
3IL-22+1-(GlcNAc2Man3Fuc)-1-(GlcNAc2Man3)218684.618680
4IL-22+2-(GlcNAc2Man3Fuc)218832.918827
5IL-22+1-(GlcNAc2Man3Fuc)117792.117788
6IL-22016749.016749
1296Xu et al. Insect-cell-derived IL-22Acta Cryst.(2004).D60,1295±1298
Acta Cryst.(2004).D 60,1295±1298Xu et al.
Insect-cell-derived IL-22
1297
162Da)and one fucose moiety (Fuc,147Da)(Manneberg et al.,1994).Analysis of the MS data reveals that heterogeneity in the glycans attached to IL-22Dm mostly arises from the presence or absence of a fucose moiety.For example,IL-22Dm variants with three glycans attached to the N X S/T sites (peak 1and 2)contain two hexasaccharides (GlcNAc 2Man 3Fuc)and one pentasaccharide (GlcNAc 2Man 3)or one hexasaccharide and two pentasaccharides (Table 1).The combination of SDS±PAGE and MS data clearly show that IL-22Dm glycosylation variants containing three or two glycans (peaks 1±4)are t
he predominant forms of IL-22expressed lanogaster S2cells.Although our studies clearly de®ne the number and type of glycans attached to IL-22Dm,we cannot identify which glycan is attached at a speci®c N X S/T site.
3.2.Crystallization
Initial crystal screening experiments were performed using IL-22Dm puri®ed by nickel-af®nity and anion-exchange chroma-tography (Fig.1,lane a ;Fig.2,lane S ).As described above,SDS±PAGE and MS analysis revealed that the protein prepara-tion was heterogeneous owing to the presence of N-linked glycosylation attached
to each of the three N X S/T glycosylation sites.Despite the chemical heterogeneity of IL-22Dm,crystallization screens performed with the Crystal Screen I kit (Hampton Research)at 298and 277K yielded small crystals in conditions No.9and 40at 298K.Condition No.9contained $5Â$15m m needles and $30m m crystalline aggregates grew in condition No.40.Re®nement of the initial conditions revealed that 0.7m M cetyltrimethylammonium bromide (CTAB)improved the size of the crystals ($0.1mm),but the morphology actually became worse.Increasing the temperature of the crystal-lization experiments to 310K resulted in a considerable improvement in morphology,but the size
of the crystals was limited to approximately 50m m on an edge and further optimization was not successful.
Marked enhancement in crystal size was only accomplished by additional puri®cation of the IL-22Dm glycosylation variants into the four ion-exchange chromatography fractions (F 6±F 9)shown in Fig.2.This was accomplished by dialysis of IL-22Dm into 20m M NaCl,20m M PIPES pH 6.5and subsequent loading of the material onto a 1.66ml POROS HS20column at a ¯ow rate of 2ml min À1.Glycosylation variants were eluted with a 1M NaCl gradient in 15column volumes.Interestingly,the largest
single crystals suitable for data collection could only be grown from IL-22Dm variants containing three or two N-linked glycans (fractions 6and 7).IL-22Dm fractions containing one or two glycans (F 8and F 9)did not yield crystals suitable for X-ray diffraction experiments (Fig.3).Although fraction F 9contains some IL-22Dm with three and two N-linked glycans,as found in fractions F 6and F 7,this material is not amenable to crystallization.Possible reasons for this are the contamination of the frac-tions with IL-22glycosylation variants containing only one glycan or a shorter or different chemical composition of the glycans present on each asparagine residue as suggested by MS analysis.The best crys-tals of IL-22Dm were grown from the early eluting fraction 7(Figs.2and 3).
3.3.X-ray diffraction
Crystals of IL-22Dm were ¯ash-cooled for low-temperature data collection in a nitrogen stream (100K)following exchange of the drop solution with a solution of 0.1M trisodium citrate pH 6.0,0.2M ammonium acetate and 29%PEG 4000.The cryo-solution was pre-warmed to 310K prior to transferring the crystal from the crystal-lization drop.Diffraction data were collected at the South Eastern Region Collaborative Access Team (SER-CAT)beamline ID-22at the Advanced Photon Source,Argonne National Laboratory.Data were collected on a MAR 165CCD detector
using a wavelength of 1.25A
Ê,an oscillation range of 1 and an exposure time of 1s,
resulting in a complete 2.6A
Êresolution data set.Re¯ection intensities were indexed,integrated and scaled using HKL 2000(Otwinowski &Minor,1997).The HKL 2000indexing routine combined with the analysis of systematic absences identi®ed the space group as P 21,with unit-cell parameters
a =64.88,
b =62.23,
c =139.52A
Ê, =91.35
.Figure 3
Crystals of IL-22Dm obtained with cation-exchange fractions 6±9shown in Fig.2.The measurement bar corresponds to 0.1mm.
Table 2
Data collection.
Values in parentheses are for the highest resolution shell.Wavelength (A Ê) 1.255
Resolution (A Ê)
50±2.6(2.69±2.60)No.observations
108300
No.unique observations 32705(2720)Redundancy
3.3(2.8)Completeness (%)9
4.5(79.5)I /'(I )
21.3(2.1)R merge (%)0.051(0.344)Space group
P 21Unit-cell parameters
a (A Ê)64.88
b (A Ê)62.23
c (A Ê)139.52 ( )
91.35
Data-collection statistics are shown in Table2.
Since it is assumed that IL-22Dm crystals contain IL-22molecules that contain two or three N-linked glycans,an average mole-cular weight of19204Da was estimated from peaks1±4in Table1.Calculations using this molecular weight resulted in three reasonable solvent estimates of66%(V M= 3.7AÊ3DaÀ1),58%(V M=2.9AÊ3DaÀ1)and 49%(V M=2.4AÊ3DaÀ1),corresponding to four,®ve or six molecules in the asymmetric unit,respectively(Matthews,1968).To distinguish between these choices,a self-rotation function(SRF)was performed with AMoR e using data in the resolution range 20±4AÊ(Nav
aza,1994).Excluding the origin peak,the three largest peaks in the SRF map correspond to three distinct non-crystallo-graphic twofold axes.A non-crystallographic twofold axis was previously observed in the asymmetric unit of IL-22Ec crystals(Nagem et al.,2002).Together,the data suggests IL-22Dm crystals contain six molecules of IL-22comprised of three non-crystallo-graphically related IL-22dimers that pack in a yet unknown manner.Molecular replace-ment and re®nement of the structure to
con®rm this hypothesis and elucidate the
role of the N-linked carbohydrate in IL-22
structure and function are currently under
way.
This work is supported by the NIH(grant
AI47300).Use of the Advanced Photon
Source was supported by the US Depart-
ment of Energy,Of®ce of Science,Of®ce of
Basic Energy Sciences under Contract No.
W-31-109-Eng-38.We thank Landon Wilson
for help with mass spectrometry and
acknowledge equipment and partial oper-
ating costs are provided by grants
S10RR11329and P30CA-13148.
References
Aggarwal,S.,Xie,M.H.,Maruoka,M.,Foster,J.&
Gurney,A.L.(2001).J.Interferon Cytokine
Res.21,1047±1053.
Dumoutier,L.,Louahed,J.&Renauld,J.C.
(2000).J.Immunol.164,1814±1819.
Dumoutier,L.,Van Roost, E.,Colau, D.&
Renauld,J.C.(2000).Proc.Natl Acad.Sci.
USA,97,10144±10149.
Fickenscher,H.,Hor,S.,Kupers,H.,Knappe,A.,
Wittmann,S.&Sticht,H.(2002).Trends
Immunol.23,89±96.
Kotenko,S.V.,Izotova,L.S.,Mirochnitchenko,
O.V.,Esterova,E.,Dickensheets,H.,Donnelly,
R.P.&Pestka,S.(2001).J.Biol.Chem.276,
2725±2732.
Logsdon,N.J.,Jones,B.C.,Josephson,K.,Cook,J.
&Walter,M.R.(2002).J.Interferon Cytokine
Res.22,1099±1112.
Manneberg,M.,Friedlein,A.,Kurth,H.,Lahm,H.
W.&Fountoulakis,M.(1994).Protein Sci.3,
30±38.
Matthews,B.W.(1968).J.Mol.Biol.33,491±497.
Nagem,R.A.,Colau,D.,Dumoutier,L.,Renauld,
J. C.,Ogata, C.&Polikarpov,I.(2002).
Structure,10,1051±1062.
Navaza,J.(1994).Acta Cryst.A50,157±163.
Otwinowski,Z.&Minor,W.(1997).Methods
Enzymol.276,307±326.
Pestka,S.,Krause,C.D.,Sarkar,D.,Walter,M.R.,
Shi,Y.&Fisher,P.B.(2004).Annu.Rev.
Immunol.22,929±979.
Walter,M.R.(2002).Immunol.Res.26,303±308.
Walter,M.R.&Nagabhushan,T.L.(1995).
Biochemistry,34,12118±12125.
Wei,C.C.,Ho,T.W.,Liang,W.G.,Chen,G.Y.&
Chang,M.S.(2003).Genes Immun.4,204±211.
Xie,M.H.,Aggarwal,S.,Ho,W.H.,Foster,J.,
Zhang,Z.,Stinson,J.,Wood,W.I.,Goddard,
A.D.&Gurney,A.L.(2000).J.Biol.Chem.
275,31335±31339.
1298Xu et al. Insect-cell-derived IL-22Acta Cryst.(2004).D60,1295±1298
发表评论